We used the AFM as the primary imaging tool and have seen the complex form very clearly. We used various DNA lengths (90, 150, 350, 500 and 1000 bps) whose end sequences are identical to those of HIV DNA ends, as substrates. The complex formation efficiency depends on the DNA length albeit not too strongly. It is interesting that the primary product in these constructs is that two DNA molecules bind to a protein tetramer and these complexes go on to aggregate by binding among the protein oligomers, forming spider-like structures. We quantified the volumes of the bound particles in two-DNA/protein complexes and confirmed them to be consistent with bound tetramers. We also examined the integration process to "host" DNA and observe stable synaptic complexes as well as fully integrated DNA. IN also appears to affect the conformational state of DNA molecules away from their binding ends, as we observe supecoil-like structures formed by two distinct such linear molecules. We are in the process of clarifying this phenomenon which would shed light into the detailed mechanics of integration and would explain the dependence of IN binding affinity to DNA length even though IN appears to bind DNA only at its active ends. Time-course studies of complex formation strongly suggests the reactions pathway: A protein tetramer first binds to a single DNA resulting in conformational changes which strongly promote binding of the second DNA.